RETHINKING FEED-IN TARIFFS AND PRIORITY DISPATCH
FOR RENEWABLES
Mark Andor ● Kai Flinkerbusch1 ● Matthias Janssen ● Björn Liebau ● Magnus Wobben
We gratefully thank FORMAET for supporting our research.
Münster, April 2010
Abstract
Within the German System of feed-in tariffs for renewable electricity supply (RES) producers of
renewable electricity also have the privilege of priority dispatch. Depending on the design of the
tariff this is either a physical priority dispatch (“guaranteed grid access”) or a financial priority
(“bonus payments”). In either case suppliers of renewable energy sources are inclined to deliver
energy even when the cost of production exceeds the market price, i.e. the electricity’s value. We
suggest to remove the priority dispatch and to modify the design of feed-in tariffs in such a way that
RES suppliers receive a payment for their potential supply in cases where the price of electricity
drops below their marginal costs. Thereby, renewable electricity producers will suffer no drawbacks
but social welfare increases.
1
Lehrstuhl für Volkswirtschaftstheorie, Westfälische Wilhems-Universität Münster
Universitätsstr. 14 – 16, 48143 Münster
Tel.: 0049251 – 83 22842
Fax: 0049251 – 83 28317
e-mail: [email protected]
1. Introduction
Promoting the use of renewable energy sources (RES) is a central goal of the European Union, which
was recently corroborated in directive 2009/28/EC. RES can be supported with either price- or
quantity-based policies.2 Feed-in tariffs, a price-based policy, are the most commonly used support
schemes3 and are also considered to be “the most efficient and effective support schemes for
promoting renewable electricity”4. The fundamental characteristic of feed-in tariffs is a guaranteed
price for the production of electricity that covers the long-term marginal costs of RES.5 Such a
support scheme has led to a massive development of RES in Germany and several other countries.6 In
this paper we do not challenge the general efficiency or effectiveness of feed-in tariffs but we show a
shortcoming of such support schemes that comes along with the linkage of the tariff and the “priority
feed-in”, which implies that the tariffs are paid for the actual production of electricity. Once the price
for electricity drops below the short-term marginal costs of RES it is not socially optimal to produce
electricity from RES. Thinking about solar- and wind power plants this may not be of importance in
times of positive prices for electricity. However, whenever negative prices occur this is of particular
relevance. Any institutional or financial priority feed-in of electricity from RES leads to a loss of
welfare. Thus, we argue for a modification of the design of feed-in tariffs without endangering the
expansion of RES. For this purpose, the remainder of this paper is as follows. Chapter 2 briefly
outlines the need for potentially unlimited negative prices. In chapter 3 we assess the loss of welfare
of the current German RES support scheme with a comparative-static analysis. In chapter 4 we
develop a simple modification of the design of feed-in tariffs with the constraint that renewable
electricity producers will suffer no drawbacks but social welfare increases. Chapter 5 concludes.
2. Negative Prices and short term intertemporal marginal costs
A price indicates scarcity and brings supply and demand to equilibrium. With electricity, there are
two peculiarities: Peak loading and non-storability. The former leads to the need for different
technologies: base- and peakload power plants. If demand were constant, a single (the cheapest)
technology would suffice. Baseload power plants have high fixed costs and low variable costs and
they are exceptionally inflexible. In an intertemporal context an inflexible power plant does not want
to change its workload as this causes additional fuel costs, attrition and, most importantly, a loss of
benefit. Whenever an inflexible plant is shut down, it needs several hours to get back to its full
workload. When prices are high in these hours, it may have been advantageous to not shut the plant
down at all. These opportunity costs must be kept in mind. In brief we call them “short term
intertemporal marginal costs of production”. These are negative for certain kinds of power plants
(Ockenfels/Grimm/Zoettl 2008).
Non-storability implies that supply must be equal to demand at all times. When there is a situation
with very low demand (for only a short time), then power plants with negative intertemporal
2
For a survey of different support schemes see IEA 2008.
See REN212009, p.18.
4
See European Commission 2008, p.3.
5
For a detailed review of several Feed-in tariff schemes see Couture/Gagnon 2010, pp.955-965.
6
See Butler/Neuhoff 2008, p. 1859 or Sawin 2004, p. 8.
3
2
marginal costs want to pay for not shutting down. So there may be situations with negative prices
even without the integration of RES. The bottom line is that negative prices are not a phenomenon
only linked to the expansion of RES supply.7
3. The inefficiency of the priority dispatch
Feed-in tariffs typically include a fixed tariff paid for actually produced electricity as well as a
guaranteed grid access that we will denote “priority dispatch”. In terms of the priority dispatch,
renewable electricity sources (RES) are drawn prior to conventional sources. Suppliers of RES are paid
a fixed feed-in tariff. The feed-in tariff is supposed to cover the long-term marginal costs, which
means capital costs are included. Capital costs have no influence on the short-term power plant
dispatch and are excluded from the following considerations. We will focus on short term
intertemporal marginal costs. In the case of RES like wind energy or photovoltaic, these costs hardly
differ from short term static marginal costs due to very low costs of output adaption. However, in the
case of biomass power plants, they can very well differ from the static marginal costs. Hence the
term “short term intertemporal marginal costs” will shortly be addressed with “marginal costs”.
Using a framework of comparative-statics we show that a welfare loss is incurred if priority dispatch
for a uniform technology of renewable electricity generation is granted.
Case 1: Price above marginal costs
Assume that all external costs and benefits are internalized. The supply function of the conventional
suppliers is linearly increasing. The uniform RES-technology has constant marginal costs cRES. Demand
is completely inelastic up to the prohibitive price. The price is p0, with p0 > cRES. In this case the supply
of renewable electricity is economic. In other words the RES-technology could produce in a market
without priority dispatch. In the long term p0 will not suffice to cover fixed costs. Hence a tariff is paid
which lies well above market prices.
In this case however the priority dispatch does not change the logic of the power plant dispatch
according to marginal costs along the merit order. This implies that under these circumstances the
RES feed-in does not cause allocative inefficiencies and thus does not cause welfare losses. Because
of the feed-in tariff there is, however, a redistribution of economic rents from consumers to RESsuppliers. This is politically intended and socially accepted.
Case 2a: Price below marginal costs and fixed feed-in tariff
7
Consequently, from an economic point of view it should become clear that a price floor for electricity (which
is widely discussed at present) cannot be an efficient way to cope with too much RES supply as the price works
as a scarcity indicator even when it becomes negative: Inflexible power plants need the signal of negative
prices to shut down. With a price floor, allocation of supply would be done by e.g. “pro rata”-allocation which
implies strong welfare losses, see Viehmann/Sämisch 2009.
3
Now assume that due to, say, low demand the market price lies below RES’ marginal costs. RES
would not produce in this case. This is shown in figure 1.
p
Demand
CRES
A
Supply
p0
B
Q
Quantity
Figure 1: Market result without priority feed-in.
The market price is p0. The consumers’ surplus is represented by area A. Area B shows the producers’
surplus. Now assume that the priority feed-in drives the RES suppliers to produce. The new dispatch
and the allocation and distribution effects are shown in figure 2.
By means of the priority dispatch the RES supply is privileged in comparison to conventional supply
and the conventional supply is pushed to the right (figure 2a),8 while the demand function does not
change. We now have an expanded supply and unchanged demand, so a part of the conventional
supply is pushed out of the market. These suppliers lose their economic surplus, represented by area
C in figure 2b. This is a welfare loss, resulting from the crowding out of economic supply. Hence the
priority dispatch leads to allocative inefficiency. In addition, marginal costs of the RES technology cRES
lie above the value of electricity, which is p0. Now as electricity is produced at a higher price than its
economic value, resources are spoiled. Thus we have an additional welfare loss, represented by area
D. The overall welfare loss, which is shown by areas C and D, originates from the fact that due to the
priority dispatch an (in this particular situation) inefficient technology is pushed into the market.
Beside these allocative effects there accrue distributional effects: Because of the expanded supply
price falls from p0 to p1.9 This implies that consumers have a surplus, represented by areas E and F in
figure 2b. Area E was a producers’ surplus beforehand – due to the decreased price there is a
redistribution in favor of the consumers. RES-producers would lose areas D and F as their marginal
costs lie above the market price. Now let us assume that they get a fixed feed-in tariff covering
exactly their marginal costs and that the tariff is paid by the consumers. In this case, the consumers
will lose areas D and F. Altogether, the new consumers surplus is A + E + F – F – D = A + E – D.
Whether the consumers are winners or losers of the priority feed-in depends on the difference on
cRES and p0 and on the quantity of produced RES-electricity. When the feed-in tariff lies above the
marginal cost of RES production there is an additional redistribution from consumers to RES
8
By using a comparative-static framework like we do not all arising effects can be incorporated. Of course the
priority feed-in changes the market rules and thus the calculus of the suppliers. In the long run the supply
function will change which may alter the identified effects. See Nicolosi/Fürsch 2009. Insofar an additional
analysis to explain the reaction of the suppliers seems appropriate.
9
This effect is often addressed with „merit order-effect of RES“. Economic implications of this effect are e.g.
covered by Sensfuß/Ragwitz/Genoese 2008.
4
producers. The conventional producers’ surplus is B – C – E. The conventional producers are
unambiguously the losers of the priority feed-in as they are partly pushed out of the market and
suffer the decreased market price.
p
p
Demand
cRES
cRES
A
SupplyRES
p0
p1
p1
B
QRES
p0
Demand
A’
D
F
C
E
SupplyRES
B’
Q
Quantity
QRES
Q
Quantity
Figure 2: The changed result with priority feed-in.
Case 2b: Price below marginal costs and bonus payment
A widely discussed alternative to the fixed feed-in tariff in combination with priority dispatch is a
model of direct marketing of RES in combination with a bonus payment.
In Germany, direct marketing has been introduced with § 17 EEG. As the revenues from the
electricity market may not cover long term marginal costs, this possibility however is scarcely used.10
A model widely discussed to promote direct marketing is the “bonus model” suggested by
Sensfuß/Ragwitz 2009. When a RES-supplier chooses direct marketing he gets a bonus payment for
each unit of produced and sold electricity. The bonus consists of a variable market premium as well
as a fixed premium. We are skeptical about the fact that the bonus payment is coupled with the
physical production and sale of electricity. This implies incentives which are similar to the incentives
in the case of the priority dispatch: As RES-suppliers only receive a bonus payment when they
actually produce electricity, they are inclined to produce already at the point where the market price
is equal to (or higher than) their marginal costs reduced by the bonus payment, i.e. they are inclined
to produce electricity at market prices below their marginal cost of production which, again, leads to
welfare losses. These are shown in figure 3.
10
See BDEW 2009 for data concerning direct marketing in 2009.
5
BONUS
p
Demand
CRES
A’
p0
D
C
F
p1
Supply
B
Q
Quantity
Figure 3: Bonus payment
The figure is similar to figure 2, albeit it shows the bonus payment and not the fixed feed-in tariff. By
means of the bonus payment the RES-suppliers can produce profitably. Market price drops to p1,
leading once more to a crowding out of conventional suppliers. A welfare loss accrues which is
identical to the loss in the case of a fixed feed-in tariff, shown by area C. Also, a loss represented by
area D accrues which is due to RES’ high marginal cost of production. (cEE > p0). We can think of
situations in which the market price is at a level where RES production is not profitable even with a
given bonus payment. In this case RES suppliers will not supply electricity and no welfare loss
accrues.
The bonus payment comes without physical priority dispatch, but it establishes a kind of “financial
priority” with similar effects on the merit order and accordingly similar welfare losses. A market price
driven dispatch of RES is not achieved.
Case 3: The example of wind energy supply
We now apply the framework developed to wind energy. The economic mechanisms are the same as
in case 2. The only difference is that we now assume marginal costs of production that are equal to
zero and that the (fixed) feed-in tariff lies considerably above these marginal costs. Figure 4 shows
this case.
Let´s assume a situation in which a low level of demand and the marginal cost situation of
conventional plant lead to a market price p0 below zero. Wind energy cannot supply profitably in this
situation. By means of the priority dispatch (or by a sufficiently high bonus payment) the wind energy
suppliers are still inclined to produce electricity. By crowding out conventional suppliers we have
again a welfare loss shown by area C. Because the price p0 lies above marginal costs of production we
also have a welfare loss shown by area D. As the market price falls to p1, consumers gain a surplus
represented by areas D, E and F. Conventional suppliers lose areas C and E. Their surplus is finally B’
only.
When the feed-in tariff is V, wind energy suppliers capture part of the consumers’ surplus. This
corresponds to areas D, G and F. The aggregate effect is again unknown as it depends on the supplied
quantity of wind energy and on the tariff. Without feed in tariffs and priority dispatch, wind energy
6
producers would not supply electricity when market prices fall below zero and no welfare loss would
accrue.
p
Demand
V
p
Demand
V
A
c=
RES0
p0
B
G
Supply
Q
c=
RES0
p0
Quantity
Supply
Q
D
F
p1
A’
Quantity
E
C
B’
Figure 4: The case of wind energy
What about external effects?
So far the analysis does not take account of external effects. We will now show that external effects
legitimate the promotion of RES, but not priority dispatch for RES.11
From a political perspective the promotion of RES shall lead to learning effects and thus to
technological advantages: RES shall become competitive over time. In addition a domestic industry
shall be established. Learning effects are achieved by research activities as well by site construction.
(Junginger/Faaij/Turkenberg 2005). The actual supply of electricity however does not lead to learning
effects. Consequently, to achieve technological advantages the promotion to build and operate
capacities should be established. But there is no benefit in maximizing the utilization of these
capacities with regard to the learning curve.
A different case concerns the emission of carbon dioxide. RES production leads to reduced emissions
of carbon dioxide by substituting conventional electricity supply. Consequently, positive external
effects are gained with each produced unit of RES electricity. However, since 2005 there is the
European emissions trading system (EU ETS) which internalizes the external effects of carbon
dioxide.12
Thereby the comparative disadvantage of RES production is reduced. When RES’ marginal costs of
production are above the market price for electricity including the price for carbon dioxide emissions,
11
Whether a promotion of actually produced RES electricity makes sense at all is not our concern. See for a
discussion e.g. Klessmann/Nabe/Burges 2008 and Lienert/Wissen 2006.
12
See Janssen/Ströbele/Wobben 2007.
7
electricity production by conventional sources should be preferred. Thus priority dispatch does not
imply a benefit for the climate.13
From a resource political view RES imply a higher level of security of supply by reducing the
dependency of imports. The social utility originates from the option to use RES when supply is scarce
and thus from the installed capacity. The bottom line is that the construction of capacity should be
promoted and not the actual supply. The priority dispatch does not lead to additional benefits.
4. Political implications
The question is what kind of support scheme provides the right incentives for RES suppliers. We have
shown that by abandoning the priority dispatch welfare gains are achieved. Without this priority in
general two cases are possible:
1. Market price is above marginal costs of RES production. In this case, RES suppliers will
produce electricity and be paid the feed-in tariff for their actual production. There will be no
welfare loss but there will be a redistribution of wealth in favor of the RES suppliers.
2. Market price is below marginal costs of RES production. In this case, RES suppliers should not
produce. In order not to change their investment decisions they would have to be paid for
their losses, i.e. they should receive the feed-in tariff for their potential production, albeit
reduced by their marginal costs of production. The deduction is necessary as otherwise RES
suppliers with positive costs of production (e.g. biomass) would have an incentive NOT to
produce in order to receive the complete tariff AND save their production costs. By not
producing and paying for the potential production welfare losses are avoided, but the
payment to RES suppliers remains unchanged to the case of priority dispatch.
For the accomplishment of our suggestions the marginal (intertemporal) costs of production of all
technologies have to be known by the entity selling the RES production. In the case of biomass power
plants the development of such a benchmark could be somewhat more complicated. In the case of
solar power plants and wind energy the investigation should be easy.14 The selling entity will then
place the bids for the forecasted RES production on the basis of the marginal costs-benchmark in the
spot market and has to make sure that production is switched off in case the market clears at a lower
price. This means that a given RES supplier will not produce electricity when the spot price lies below
her specific marginal cost benchmark. Thus welfare losses are avoided.
However, to achieve these efficiency gains investments are required as not all RES technologies can
currently be monitored and adjusted in their output. Such investments however will make RES
production “smarter” which might even be a prerequisite in order to increase the share of RES in line
with political objectives.
13
When RES promotion and emissions trading are not coordinated by adapting the emissions cap, the priority
feed-in does even constrain the efficiency of climate policy. Without an adaption of the cap the priority feed-in
leads to reduced emissions prices leading to comparative advantages for those power plants with the highest
emissions. See Böhringer/Rosendahl 2009, Kemfert/Diekmann 2009 and Kemfert/Traber 2009.
14
One could assume them to be very close to zero.
8
5. Summary
We have shown that the price for electricity is an essential scarcity indicator, no matter if positive or
negative. The reason why negative prices are observed more and more often is the growth of
renewable energy sources which enjoy the benefits of a feed in tariff.. However, the priority feed-in
implies that RES suppliers are paid for their actual electricity generation, thereby causing them to
supply independent from the price of electricity. This situation leads to welfare losses. Paying a
bonus to cover fixed costs –as is the case in the model proposed by Sesfuß/Ragwitz- does not
eliminate the welfare losses as long as payment is tied to actual production.
The losses could be avoided by abandoning the priority dispatch and paying the RES suppliers for
their potential supply in cases where the price, or value, of electricity drops below their marginal
costs. The payment has to be reduced by the suppliers’ marginal costs of production which are
avoided when no electricity is actually produced. This design would lead to significant welfare gains
and could pave the way for widespread direct marketing. Future research should be directed to
define the details of such a model to promote renewable energy sources in a more efficient way,
tailored to be in line with competitive and open markets.
9
References
BDEW (2009). Angaben zur Direktvermarktung nach § 17 EEG in den einzelnen Monaten des Jahres
2009. Stand vom 10. November 2009. URL http://www.bdew.de/bdew.nsf/id/DE_EEGMonatsprognosen?open.
Böhringer C, Rosendahl K (2009). Green Serves the Dirtiest. On the Interaction between Black and
Green Quotas. Statistics Norway, Research Department, Discussion Paper No. 581.
Butler, L., Neuhoff, K. (2008). Comparison of feed-in tariff, quota and auction mechanisms
to support wind power development. Renewable Energy, 33, 1854–1867.
Couture, Toby and Gagnon, Yves (2010). An analysis of feed-in tariff remuneration models:
Implications for renewable energy investment, Energy Policy, 38, 2, 955-965.
European Commission (2008). Commission Staff Working Document, Brussels, 57, 23
January 2008. [http://ec.europa.eu/
energy/climate_actions/doc/2008_res_working_document_en.pdf].
International Energy Agency (IEA) (2008). Deploying Renewables: Principles for Effective
Policies, Paris.
Janssen M, Ströbele W, Wobben M (2007). Electricity pricing subject to a CO2-emissions trading
scheme. Zeitschrift für Energiewirtschaft, 3, 171-181.
Junginger M, Faaij A, Turkenburg W (2005). Global experience curves for wind farms, Energy Policy,
2, 133-150.
Kemfert C, Diekmann J (2009). Förderung Erneuerbarer Energien und Emissionshandel: wir brauchen
beides. Wochenbericht des DIW Berlin, 11, 169-174.
Kemfert C, Traber T (2009). Impacts of the German Support for Renewable Energy on Electricity
Prices, Emissions, and Firms. The Energy Journal, 3, 155-178.
Klessmann C, Nabe C, Burges K (2008). Pros and cons of exposing renewables to electricity market
risks – A comparison of the market integration approaches in Germany, Spain, and the UK.
Energy Policy, 10, 3646 – 3661.
Lienert, M, Wissen, R (2006). Bewertung von Fördersystemen für Erneuerbare Energien: Eine
kritische Analyse der aktuell geführten Diskussion. Zeitschrift für Energiewirtschaft, 30, 133 –
140.
McDonald A, Schrattenholzer L (2001). Learning rates for energy technologies. Energy Policy, 4, 255
– 261.
Nicolosi M, Fürsch M (2009). The Impact of an increasing share of RES-E on the Conventional Power
Market – The Example of Germany. Zeitschrift für Energiewirtschaft, 3, 246 – 254.
10
Ockenfels A, Grimm V, Zoettl G (2008). Strommarktdesign – Preisbildungsmechanismen im
Auktionsverfahren für Stromstundenkontrakte an der EEX. Gutachten im Auftrag der
European Energy Exchange AG zur Vorlage an die Sächsische Börsenaufsicht.
REN21 (2009). Renewables Global Status Report: 2009 Update. REN21 Secretariat, Paris.
Sawin J. (2004). National policy instruments: policy lessons for the advancement and
diffusion of renewable energy technologies around the world. Thematic Background
Paper. In: Proceedings of the International Conference for Renewables, Bonn, 2004.
Sensfuß F, Ragwitz M (2009). Entwicklung eines Fördersystems für die Vermarktung Erneuerbarer
Stromerzeugung. 6. Internationale Energiewirtschaftstagung an der TU Wien, 2009.
Sensfuß F, Ragwitz M, Genoese M (2008). The merit-order effect: A detailed analysis of the price
effect of renewable electricity generation on spot market prices in Germany. Energy Policy, 8,
3086 – 3094.
Viehmann J, Sämisch H (2009). Windenergieintegration
Energiewirtschaftliche Tagesfragen, 59, 49 – 51.
bei
negativen
Strompreisen.
11